The government regulates many harmful things — why isn’t AI one of them?
Illustration by Sage Blackwell.
In B.C., you cannot consume alcohol, cannabis, or tobacco if you are under the age of 19. You cannot drive without a license, which you cannot obtain if you are under the age of 16. You cannot possess or use a firearm if you don’t have a licence. The list goes on.
So why is it that you can utilize generative artificial intelligence (AI) with almost entirely free rein?
I know what you’re thinking — those things I listed before have the capacity to cause great harm to individuals and the public. Generative AI writes my emails and answers my questions, it doesn’t increase my chances of lung cancer, or impair my driving. So what’s the link between them?
Many things have negative immediate or long-term effects, pose personal or public safety risks, and need to be used responsibly. Most of us can agree that one important function of a democratic government is to assess the risk of harm that accompanies any one thing, then implement and enforce laws around it to protect citizens.
We see this every day, and generally accept it as in the interest of the greater good; we acknowledge the harm that it protects against. The question is: how does the government assess and define harm, and how should they?
On Nov. 30, 2022, OpenAI rolled out the AI program ChatGPT for public use, and within only five days, the program had over 1 million users. At the time, Canada had no laws in place to govern the development and operation of generative AI. Today, almost four years later, Canada still has no meaningful legal or policy framework to address the harms of generative AI.
While the Canadian government has made meek attempts at policy, nothing has stuck. OpenAI itself has recently implemented an age minimum of 13 for users, and requires that users aged 13-18 obtain parental consent to use the program. However, it remains unclear exactly how age and consent will be verified, and how these rules will be enforced.
Excessive use of alcohol and marijuana by youth can interfere with brain development. According to a recent MIT study, excessive reliance on generative AI may contribute to cognitive atrophy and the shrinking of critical thinking skills.
Firearms and driving can both cause serious injury or death if used irresponsibly. Two recent lawsuits against OpenAI allege ChatGPT’s connection to both the Florida State University (FSU) shooting in April 2025 and the Tumbler Ridge shooting in February 2026.
Families of Tumbler Ridge victims are suing OpenAI, alleging that the 18-year-old shooter conversed with ChatGPT about gun violence scenarios and had been flagged. The lawsuit alleges safety team members recommended contacting authorities, but OpenAI’s leadership overruled them.
Similarly, a family member of an FSU shooting victim has filed a federal lawsuit against OpenAI, alleging that the shooter shared images of the firearms he’d acquired with the chatbot and it explained how to use them. The lawsuit alleges the chatbot told him that it was likelier for a shooting to gain national attention if children are involved.
In 2025, 23-year-old Zane Shamblin committed suicide after confiding in the chatbot, which allegedly encouraged him repeatedly to do so. Later that same year, 16-year-old Adam Raine killed himself after months of alleged encouragement from ChatGPT.
Raine’s family claims the chatbot discussed methods of suicide with him, and even offered to help him write a suicide note to his parents.
These heartbreaking tragedies illustrate the inherent safety risks that accompany unregulated generative AI. Each tragedy alone should have been enough for a call-to-action.
AI image integration poses a completely different but equally concerning threat. Beyond the ease with which it aids misinformation to spread online, it can and has been used in targeted attacks on women.
X’s (Formerly Twitter) built in generative AI tool, Grok, was updated in January to have photo-editing capabilities. Anyone with an X account had access to this tool. Within only days requests flooded in to depict users in small bikinis, degrading poses, covering them in bruises or “doughnut glaze,” and placing gags in their mouths, among others.
Beyond the horrifying realities of how generative AI is being utilized exist various other categories of consequences.
For example, the environmental impacts of generative AI are astronomical. A single prompt can use up to ten times as much electricity as a single Google search — and image generation requires even more power. Data centers consume mass amounts of energy and water, even causing water depletion in some already water-scarce areas.
The list goes on.
There is no denying that generative AI is an incredible innovation that can be used for great progression in many fields of research. There is also no denying that we are past the point of no return — generative AI is a part of our world and we are not going back.
At this point, the best we can do is to demand better regulation. OpenAI should be mandated to monitor conversations and report safety violations immediately. Image generation should not be a tool readily available to anyone who wants to use it. Criminal record checks should, under some circumstances, be a requirement before creating an account.
While it took government years to understand the harms of alcohol, tobacco, marijuana, and more, and it has only been three and a half years since the public release of ChatGPT, the short-term effects are immediately in front of us, and the long-term effects can be easily predicted. The consequences could not be more clear and there is no time to wait. The time to take action is now.
